International Business Machines Corporation

Spyros Kotoulas

Marco Luca Sbodio

Aris Gkoulalas-Divanis

Anastasios Kementsietsidis

Pierpaolo Tommasi

Vanessa Lopez Garcia

Achille Belly Fokoue-Nkoutche

Akihiro Kishimoto

Martin Stephenson

Natalia Mulligan

Muhammad Mustafa Rafique

Natasha Mulligan

Vanessa Lopez

Joao H Bettencourt Da Silva

Elizabeth Daly

Adi Ionel Botea

Radu Marinescu

Theodora Brisimi

Valentina Rho

Nagesh Yadav

Caroline A O'Connor

Nuno Lopes

Daniel Rejniak

James Shannon

Konstantin Katrinis

Mark B Hughes

Veli Bicer

Embodiments for generating personalized advice to a user by a processor. A health state of a user may be learned from feedback information collected from a plurality of data sources for providing one or more customized communications. One or more customized communications may be provided to a user to alter one or more activities of the user so as to avoid one or more possible negative impacts upon the health state of the user.

Cognitive health state learning and customized advice generation

Various embodiments are provided for learning policy explanations in a computing environment by a processor. One or more explanations may be provided that justify validity or invalidity of a claim based on one or more rules extracted from one or more segments of text data of a policy data source using a machine learning operation.

Learning policy explanations

Techniques are described for extracting structured information from unstructured documents based on validation of the structured information as applied to a domain problem associated with the unstructured documents. In one embodiment, a computer program product for automated information extraction is provided. The computer program product comprising a computer readable storage medium having program executable by a processing component to cause the processing component to extract structured candidate interpretations of a rule from unstructured information that defines a plurality of rules intended to control operations of a system, and determine measures of validity of the structured candidate interpretations based on application of the candidate policy interpretations to historical operational data for the system that represents operations performed by the system. The measures of validity respectively can represent degrees to which the historical operational data reflects operation of the system in accordance with the structured candidate interpretations.

Extracting structured information from unstructured data using domain problem application validation

A capability to identify and visualize relationships and commonalities amongst record entities is provided. A plurality of entities are extracted from one or more records. Each extracted entity is associated with a respective feature vector within a vector space of a feature matrix. The feature vectors are distributed within the feature matrix based on semantic relationships amongst the entities of a corpus. Multidimensional coordinates within a dimensionally-reduced vector space of the feature matrix are generated for each extracted entity. One or more cells of a cellular presentation of the feature matrix are identified such that each identified cell represents one or more respective extracted entities. Each cell represents (i) a respective range of multidimensional coordinates within the dimensionally-reduced vector space of the feature matrix and (ii) one or more feature vectors of the plurality of feature vectors within the feature matrix.

Identifying and visualizing relationships and commonalities amongst record entities

A method for unsupervised transformation of denormalized data into RDF data includes summarizing an existing resource description framework (RDF) file wherein a data summary file is produced, wherein said RDF file includes a plurality of nodes, property labels, and literals and is represented as a directed graph, and said data summary file is a graph that includes a single instance of each node-property-literal triple in the RDF file; analyzing a plurality of denormalized records wherein a record summary is produced, wherein said record summary is represented as a directed graph with a root node corresponding to each record connected to a node for each field in the record; determining one or more transformation patterns that match elements of the record summary to elements of the data summary; and transforming the plurality of denormalized records into a new RDF file by executing one of the one or more transformation patterns.

Unsupervised method for enriching RDF data sources from denormalized data

A system, method and computer program product for augmenting, in real-time, queries/questionnaires administered to a subject by coupling the physiological response of the subject for a given question-answer pair. Further, a method/system augments/annotates answers to questionnaires/surveys/etc. with automatically detected physiological or emotional reactions of the person answering the question/survey/etc. In an embodiment, methods are implemented to: create an augmented representation of question-answer-emotion (i.e., annotating questions/answers from assessments to physiological reactions); and consider the augmented representation and the user profile to suggest actions based on such associations. In one aspect, the augmented/annotated answers are used to suggest/recommend actions to the person conducting the questionnaire/survey, where actions include: asking further questions to clarify a topic/problem, asking for clarification regarding a given answer, or asking for clarification regarding inconsistencies in answers or asking follow-up questions.

Augmenting questionnaires

Aspects of the present invention disclose a method, computer program product, and system for transposing ranking models. The method includes computer processors determining an initial ranking modeling framework for a record based on analyzing data from a plurality of sources. The record includes a plurality of objects and an object includes at least one feature. The method further includes computer processors determining a mapping of relationships between features included in the record. The method further includes computer processors identifying a user selection of a feature. The method further includes computer processors determining a transposition path between the user selected feature and a set of features having a determined mapped relationship to the user selected feature. The method further includes computer processors modifying ranking information associated with features included in the determined transposition path. The method further includes computer processors reordering the record based on the modified ranking information.

Transposing of ranking models

Input text containing a plurality of patient information can be annotated for annotations and extracted. One or more annotations are parsed for relevant contextual information. The one or more annotations are assigned a semantic type. The one or more annotations are visually isolated, personalized to a user profile job, and outputted. The one or more extracted annotations are subjected to natural language processing operations.

Automatic generation of personalized visually isolated text

Accessing data stored on a remote server includes determining a first address of a first server where a block of data identified by a key has been moved from a second server to the first server, the second server having a second address. A first entry of a switch flow table of a datacenter network is updated to indicate that received packets with a destination address including the second address are to be overwritten with the first address. The data center network includes a plurality of physical network switches. A second entry in the switch flow table may be configured to forward packets with the first address to a first port associated with the first server and a first set of the switches. If the second entry is not already present in the table, the second entry may be inserted into the switch flow table.

Routing packets in a data center network

Technique to determine relative relevance of pairs of nodes in a linked data graph, such as an RDF graph, including calculation of raw relevance scores (which are based upon adjacency relations between nodes after edge filtering, the incoming and outgoing degrees of nodes in the graph after the edge filtering, and the distance (or weight) values of the edge(s)) and/or 'refined relevance scores,' which take into account 'complementary nodes.'

Node relevance scoring in linked data graphs

An embodiment of the invention provides a method wherein a database query including a first constraint and additional constraint(s) are received in a first node. Data in the first node that satisfies the first constraint is identified, encoded, and sent to a second node. Encoded data is identified in a mapping table in the second node; and, one or more missing identifiers are identified that include encoded data that is not in the mapping table. The missing identifier is sent to the first node, decoded to retrieve the value of the missing identifier, and mapped to the retrieved value. The mapping of the missing identifier and the retrieved value are sent to the second node. A dictionary in the second node is queried with the retrieved value to identify an identification number for the retrieved value. The missing identifier is mapped to the identification number for the retrieved value.

Efficient distributed query execution

A method for query evaluation comprises receiving a query over a set of distributed data sources, decomposing the query into a set of sub-queries of the query, evaluating each sub-query in the set of sub-queries with respect to each data source in the set of distributed data sources, wherein evaluating comprises determining which data sources in the set of distributed data sources are capable of answering each sub-query and at what cost, computing a set of distributed plans by composing one or more of the sub-queries in one or more of the data sources, evaluating each plan in the set of distributed plans, selecting a sub-set of plans from the set of distributed plans to be executed for responding to the query, executing the selected sub-set of plans, and returning results of the query.

Systems and methods for query evaluation over distributed linked data stores

A system for processing a transaction request. A transaction request is received by a server computer that includes a user identification and an associated request token, to execute on an application on the server computer. It is determined by the server computer whether available resources on the server computer to perform the transaction are below respective threshold values. In response to determining, by the server computer, that the available resources on the server computer to perform the transaction are below the threshold value, the server computer determines whether the user identification is allowed access to the application. In response to determining that the user identification is allowed access to the application, the transaction is executed on the server computer. In response to determining that the user identification is not allowed access to the application, the transaction is rejected.

Prevention of query overloading in a server application

A method, system and computer program product for transforming a Linked Data graph into a corresponding anonymous Linked Data graph, in which semantics is preserved and links can be followed to expand the anonymous graph up to r times without breaching anonymity (i.e., anonymity under r-dereferenceability). Anonymizing a Linked Data graph under r-dereferenceability provides privacy guarantees of k-anonymity or l-diversity variants, while taking into account and preserving the rich semantics of the graph.

Growing community of inventors

Dublin, Ireland

Spyros Kotoulas